POTENCY METHOD VALIDATION C HUNG C HOW C HAN ,P H .D.

2.4.2 Accuracy

The ICH defines the accuracy of an analytical procedure as the closeness of agreement between the values that are accepted either as conventional true val- ues or an accepted reference value and the value found. Accuracy is usually reported as percent recovery by assay, using the proposed analytical proce- dure, of known amount of analyte added to the sample. The ICH also rec- ommended assessing a minimum of nine determinations over a minimum of

Figure 2.2. Linearity with correlation coefficient greater than 0.997.

eak area 8,000 P 6,000

Concentration (µg/mL)

Figure 2.3. Linearity with correlation coefficient less than 0.997.

18 POTENCY METHOD VALIDATION

three concentration levels covering the specified range (e.g., three concentra- tions/three replicates).

For a drug substance, the common method of determining accuracy is to apply the analytical procedure to the drug substance and to quantitate it against

a reference standard of known purity. For the drug product, accuracy is usually determined by application of the analytical procedure to synthetic mixtures of the drug product components or placebo dosage form to which known quantities of drug substance of known purity have been added. The range for the accuracy limit should be within the linear range. Typical accuracy of the recovery of the drug substance in the mixture is expected to be about 98 to 102%. Values of accuracy of the recovery data beyond this range need to be investigated.

2.4.3 Precision

The precision of an analytical procedure expresses the closeness of agreement (degree of scatter) between a series of measurements obtained from multiple samples of the same homogeneous sample under prescribed conditions. Preci- sion is usually investigated at three levels: repeatability, intermediate precision, and reproducibility. For simple formulation it is important that precision be deter- mined using authentic homogeneous samples. A justification will be required if a homogeneous sample is not possible and artificially prepared samples or sample solutions are used.

Repeatability (Precision). Repeatability is a measure of the precision under the same operating conditions over a short interval of time. It is sometimes referred to as intraassay precision. Two assaying options are allowed by the ICH for investigating repeatability:

1. A minimum of nine determinations covering the specified range for the procedure (e.g., three concentrations/three replicates as in the accuracy experiment), or

2. A minimum of six determinations at 100% of the test concentration. The standard deviation, relative standard deviation (coefficient of variation),

and confidence interval should be reported as required by the ICH. Tables 2.2 and 2.3 are examples of repeatability data. Table 2.2 shows good repeatability data. However, note that the data show a slight bias below 100% (all data between 97.5 and 99.1%). This may not be an issue, as the true value of the samples and the variation of the assay may be between 97.5 and 99.1%. Table 2.3 shows two sets of data for a formulation at two dose strengths that were performed using sets of six determinations at 100% test concentration. The data indicate a definite bias and high variability for the low-strength dose formulation. It may call into question the appropriateness of the low-dose samples for the validation experiment.

STRATEGIES AND VALIDATION PARAMETERS

Table 2.2. Repeatability at Different Concentration

Concentration (Nominal Concentration 75 µg/mL)

Replicate 14.8 29.6 44.5 74.5 149.1 223.7 1 97.9 98.3 97.6 98.3 98.7 98.6

Mean 98.3 98.5 98.3 98.6 98.1 98.8 % RSD

Table 2.3. Repeatability at High and Low Concen- trations

Replicate

Low Dose

High Dose 1 94.8 100.6

Mean 94.1 100.9 % RSD

Intermediate Precision. Intermediate precision is defined as the variation within the same laboratory. The extent to which intermediate precision needs to be established depends on the circumstances under which the procedure is intended to be used. Typical parameters that are investigated include day-to-day varia- tion, analyst variation, and equipment variation. Depending on the extent of the study, the use of experimental design is encouraged. Experimental design will minimize the number of experiments that need to be performed. It is important to note that the ICH allows exemption from doing intermediate precision when reproducibility is proven. It is expected that the intermediate precision should show variability that is in the same range or less than repeatability variation. The ICH recommended the reporting of standard deviation, relative standard deviation (coefficient of variation), and confidence interval of the data.

Reproducibility. Reproducibility measures the precision between laboratories as in collaborative studies. This parameter should be considered in the standardiza- tion of an analytical procedure (e.g., inclusion of procedures in pharmacopoeias

20 POTENCY METHOD VALIDATION

and method transfer between different laboratories). To validate this character- istic, similar studies need to be performed at other laboratories using the same homogeneous sample lot and the same experimental design. In the case of method transfer between two laboratories, different approaches may be taken to achieve the successful transfer of the procedure. However, the most common approach is the direct method transfer from the originating laboratory to the receiving labo- ratory. The originating laboratory is defined as the laboratory that has developed and validated the analytical method or a laboratory that has previously been cer- tified to perform the procedure and will participate in the method transfer studies. The receiving laboratory is defined as the laboratory to which the analytical pro- cedure will be transferred and that will participate in the method transfer studies. In direct method transfer it is recommended that a protocol be initiated with details of the experiments to be performed and acceptance criteria (in terms of the difference between the means of the two laboratories) for passing the method transfer. Table 2.4 gives a set of sample data where the average results obtained between two laboratories were within 0.5%.

2.4.4 Robustness

The robustness of an analytical procedure is a measure of its capacity to remain unaffected by small but deliberate variations in the analytical procedure param- eters. The robustness of the analytical procedure provides an indication of its reliability during normal use. The evaluation of robustness should be considered during development of the analytical procedure. If measurements are susceptible to variations in analytical conditions, the analytical conditions should be suitably controlled or a precautionary statement should be included in the procedure. For example, if the resolution of a critical pair of peaks was very sensitive to the percentage of organic composition in the mobile phase, that observation would have been observed during method development and should be stressed in the procedure. Common variations that are investigated for robustness include filter effect, stability of analytical solutions, extraction time during sample prepara- tion, pH variations in the mobile-phase composition, variations in mobile-phase composition, columns, temperature effect, and flow rate.

Table 2.5 shows examples of sample and standard stability performed on an analytical procedure. The two sets of data indicate that the sample and standard

Table 2.4. Results from Method Transfer between Two Laboratories

Average % Originating

Runs

12 100.7 laboratory Receiving

4 100.2 laboratory

STRATEGIES AND VALIDATION PARAMETERS

Table 2.5. Stability of Sample and Standard Solutions

Table 2.6. Effect of Filter

Replication

Unfiltered

solutions were stable for 3 and 4 days respectively. Table 2.6 gives some data on the effect of a filter on the recovery of the analytical procedure. In a filter study it is common to use the same solution and to compare a filtered solution to an unfiltered solution. For the unfiltered solution, it is common to centrifuge the sample solution and use the supernatant liquid for the analysis. The data set indicated that filter 1 would be recommended for the final analytical procedure.

2.4.5 Specificity

Specificity is the ability to assess unequivocally an analyte in the presence of components that may be expected to be present. In many publications, selectivity and specificity are often used interchangeably. However, there are debates over the use of specificity over selectivity [6]. For the purposes of this chapter, the definition of specificity will be consistent with that of the ICH.

The specificity of a test method is determined by comparing test results from an analysis of samples containing impurities, degradation products, or placebo ingredients with those obtained from an analysis of samples without impurities, degradation products, or placebo ingredients. For the purpose of a stability- indicating assay method, degradation peaks need to be resolved from the drug substance. However, they do not need to be resolved from each other.

Critical separations in chromatography should be investigated at the appro- priate level. Specificity can best be demonstrated by the resolution of two chro- mographic peaks that elute close to each other. In the potency assay, one of the peaks would be the analyte peak. Figure 2.4 illustrates the selectivity of a method to resolve known degradation peaks from the parent peak. Based on the

22 POTENCY METHOD VALIDATION

Sample solution Response (mV) 59.000

56.000 Impurity standard solution 53.000 50.000

Time (s)

Figure 2.4. Overlay chromatogram of an impurity solution with a sample solution.

experience with the analyte and the chemistry of the analyte, the scientist will

be able to identify which of the impurities may be used as the critical pair.

2.5 POTENCY METHOD REVALIDATION

There are various situations during the life cycle of a potency method that require revalidation of the method.

1. During optimization of the formulation or drug substance synthetic process, significant changes may have to be introduced into the process. As a result, to ensure that the analytical method will still be able to analyze the poten- tially different profile of the drug substance or drug product, revalidation may be necessary.

2. The method was found to be deficient in some areas, such as precision and system suitability. This is especially important as the analytical laboratory gets more experience and more information as to the degradation profile of the sample as it progresses toward submission. If a new impurity is found that makes the method deficient, this method will need to be revalidated.

3. The composition and/or the final manufacturing process of a sample ana- lyzed with the method have been modified after optimization.

4. Changes in equipment or in suppliers of critical supplies at the time of manufacturing. This is important, as critical components of the manufac- turing process have the potential to change the degradation profile of the product.

COMMON PROBLEMS AND SOLUTIONS

2.6 COMMON PROBLEMS AND SOLUTIONS

In the following pages we summarize some of the common deficiencies of potency method validation. These common problems are grouped together into categories such as HPLC instrumentation, procedural steps, and miscellaneous errors.

2.6.1 HPLC Instrumentation Errors

Qualification of Instruments. The status of the qualification of HPLC and other equipment used for the analytical procedure must always be checked. This is a common error that can lead to reanalysis of the samples if discovered earlier, or repeating the entire experimental procedure if it was discovered after expiry of the sample solutions.

Vacuum Filtering of Mobile Phase. Vacuum filtering of the mobile phase should

be avoided in a procedure that is very sensitive to the level of the organic in the mobile phase. Vacuum suction will evaporate the volatile organic portion during filtration (e.g., acetonitrile or methanol), and may lead to variation of the chromatography.

2.6.2 Procedural Errors

Expiry of Mobile Phase. Always check the expiry of mobile phase before use. This is one of the most common errors in an analytical laboratory.

Use of Ion-Pairing Reagents in Mobile Phase. It is usually recommended that if ion-pairing reagents are needed in a mobile phase, its concentration needs to

be constant during a gradient run. Changes in ion-pairing concentration during an HPLC run will increase the likelihood of chromatographic variation between runs (e.g., retention time drifts and quantitation precision).

Quantitation of Salts (e.g., Hydrochloride and Sodium Salt). The quantitative result that is reported from the analysis of salts is usually reported with reference to the base of the analyte. The scientist will need to remember to incorporate a

multiplier into the calculation to convert the salt data to the base data. Stability of Standard and Sample Solutions. Appropriate stability of the standard

and sample solutions will allow flexibility of the method to be used in a qual- ity control laboratory. For example, 4-day stability of the standard and sample solutions will allow investigation if problems arise during a weekend HPLC run.

Dilution during Sample and Standard Preparation. Minimize the number of dilutions required to give the final dilutions of the sample and standard solutions. Each dilution step will have the potential to introduce error in the procedure.

24 POTENCY METHOD VALIDATION

Range in Validation of Linearity Is Smaller Than Precision and Accuracy. This error will invalidate the precision and accuracy data since the validation did not demonstrate the linearity of the analyte for the quantitation of precision and accuracy data.

2.6.3 Miscellaneous Errors

Validation Protocol. It is highly recommended to validate an analytical proce- dure using some form of validation protocol. Without a validation protocol, the scientist will have a tendency to vary the experiment during the course of the validation study. Getting into the habit of creating a validation protocol will also ensure that the scientist plans before starting the experiment.

Acceptance Criteria for Validation Parameter. It is highly recommended to set acceptance criteria prior to starting validation experiments. This will provide guidance to the validating scientist on the range of acceptability of the valida- tion results.

Documentation of Observation. It is very important to document all relevant observations during the experimental procedure. Observations are the most impor- tant information that can be used if an investigation is needed. Furthermore, observations that are documented provide evidence in the event of patent chal- lenge and other court cases.

Absorbance of Analyte. It is common to devise an experimental procedure that yields an analyte absorbance value of less than 1 absorbance unit. A high absorbance value (depending on the absorptivity of the analyte) is the result of

a high concentration of the analyte. Too high a concentration of the analyte may overload the column and lead to nonlinearity.

2.7 SUMMARY OF POTENCY VALIDATION DATA

It is very useful to summarize all method validation data into a tabular format. The tabulated summary will give a quick overview of the validation data. Often, the analyst may be so involved during the actual validation work that some errors escaped detection. Table 2.7 is an example of how data can be recorded.

Table 2.7. Sample Validation Summary

ICH Validation Summary Characteristic

Validation Results Accuracy

Data Reported

The percent recovery assessed Based on determinations at using a minimum of nine

three concentration levels, determinations over a

average recovery = 101.4%, minimum of three

RSD = 0.9%. concentration levels covering

the range specified.

25 Table 2.7 ( continued )

SUMMARY OF POTENCY VALIDATION DATA

ICH Validation Summary Characteristic

Validation Results Precision

Data Reported

Repeatability The standard deviation, relative A single experiment (n = 6) standard deviation (RSD),

had a repeatability (RSD) of and confidence interval

1.1%. should be reported for each type of precision investigated.

Intermediate Based on experimental design precision

(n = 24) results from three dosage strengths, the

estimated intermediate precision is 0.9%.

Reproducibility The average potency result obtained from the receiving laboratory is within ± 0.5% of results from the originating laboratory.

Specificity Representative chromatograms The placebo peaks, process demonstrate specificity.

impurities, and degradant peaks are resolved from the peak of interest (e.g., Figure 2.4).

Linearity Data from the regression line Correlation (correlation coefficient,

coefficient = 0.9999; y -intercept, slope, residual

y -intercept = 0.0328 area sum of squares) and a plot.

unit; slope = 0.5877 [area/(µg/mL)]; residual sum

of squares = 178.96 (e.g., data plot in Figure 2.2).

Range Procedure provides an The range was confirmed as 70 acceptable degree of

to 130% of the test linearity, accuracy, and

concentration. precision when applied to samples containing analyte within or at the extremes of the specified range of procedure.

(continued overleaf )

26 POTENCY METHOD VALIDATION

Table 2.7 ( continued )

ICH Validation Summary Characteristic

Validation Results Robustness

Data Reported

In the case of liquid The factors evaluated (analyst, chromatography, typical

instrument, % ACN, and variations are: pH in a

column age) did not have mobile phase, composition

any significant effect of mobile phase, different

(p > 0.05) on the potency columns (different lots

results in the ranges studied and/or suppliers),

based on JMP analysis. The temperature, and flow rate.

method is robust for all the factors studied. The standard and sample solutions were found to be stable for 5 days (at 30 ◦ C).